Executive Summary: Society of Critical Care Medicine Clinical Practice Guidelines on Adult End-of-Life Care in the ICU

Evidence-based practice has evolved significantly since the release of the 2008 Society of Critical Care Medicine (SCCM) End-of-Life (EOL) Consensus Statement. The purpose of this 2025 guideline is to identify and disseminate evidence-based recommendations grounded in current research to better understand how to deliver appropriate and effective adult EOL care in the ICU. A multidisciplinary team approach to EOL care in the ICU is crucial with the aging of the U.S. population, due, in part, to increasing life expectancy and the burdens associated with chronic disease. Research indicates at least 20% of ICU patients currently have goals that are not medically attainable or their treatment plan does not align with their achievable goals (1,2). Thus, it is paramount that we better understand how to assist with decision-making, mitigate conflict about treatment goals, address suffering, and enhance capacity for EOL care in the ICU. This includes detailed knowledge of how EOL care fits into the delivery of high-quality ICU care. These guidelines represent new knowledge and address previously unconsidered issues such as identifying and educating substitute decision makers on their roles, education, and training in primary palliative care for all ICU team members, protocolized approaches to withdrawing life-sustaining-treatment, and interventions to educate and assist patients, surrogates, and non-ICU treatment teams with decision-making before ICU care is needed. A call for interested panel members was submitted to the SCCM membership and separately to the SCCM Ethics Committee membership in early 2021. Selection criteria included: diversity of professional expertise in clinical care and/or research in topic areas associated with EOL care in the ICU and/or clinical ethics; diversity of practice environment to include urban, suburban, and rural healthcare system representation; diversity of stage in career to include early mid- and advanced-career clinicians; diversity in personal demographics; and representation of patient/family perspectives with the intent that our panel reflect the diverse communities to which our ICU patients belong. The completely convened panel was comprised of 21 persons. The SCCM Ethics Committee developed an initial list of suggested guideline questions. The guideline panel revised those questions in Population, Intervention, Comparison, and Outcome (PICO) format. The PICO questions were organized into three domains: improving communication about EOL care in the ICU, improving provision of EOL care/symptom management for patients in the ICU, and addressing the educational needs of clinicians, patients, families and surrogate decision makers in ICU EOL care. After circulating the list to the guideline panel, teleconferences refined it to ten final PICO questions. Chairs assigned panelists to working groups, which, following GRADE (a structural framework for evaluating and grading the quality of evidence) guidance, categorized outcomes as “critical,” “important,” or “of limited importance,” incorporating insights from patients, families, and multidisciplinary ICU clinicians. Medical librarians conducted systematic searches of databases from 2000 to October 8, 2024. The methodology team performed data abstraction with a standardized spreadsheet, assessing bias using the modified Cochrane Risk of Bias tool for Randomized Controlled Trials and the modified CLARITY tool (that assesses the risk of bias in cohort studies) for observational studies. Meta-analyses for each PICO question were conducted using RevMan 5.4 (a systematic review and meta-analysis tool; Cochrane, 11-13 Cavendish Square, London, United Kingdom) where applicable. Standard GRADE practice was used to rate certainty of evidence with narrative summation of qualitative or other data not amenable to meta-analysis. Each PICO question working group reviewed evidence summaries and used an Evidence-to-Decision (ETD) framework to draft recommendations. Using the ETDs, the working groups evaluated effects, evidence certainty, resources, equity, feasibility, and acceptability in drafting recommendations for their respective PICOs. The full panel then voted on each recommendation, requiring over 80% agreement from 70% of eligible voters for consensus, consistent with SCCM requirements (Table 1). Complete methods and results are located in the supplementary materials (https://links.lww.com/CCM/H790). TABLE 1. - Summary of Recommendations PICO Question Recommendation Strength of Recommendation 1.0. Improving communication about EOL care in the ICU 1.1. Should we recommend specific interventions to enhance shared decision-making between patients, family members/surrogates, and the clinical team? (clinical team can include ICU, but also other specialty services, such as palliative medicine, surgery, oncology)? 1.1. We suggest using structured tools to facilitate shared decision-making for EOL treatment decisions in the ICU. Conditional recommendation, moderate certainty evidence Remark: While there is no single ideal tool to use, those studied in the ICU setting include communication facilitators; structured meeting plans; and paper/web-based decision aids. 1.2. Should we recommend specific interventions to ensure procedural due process for substitute decision-making on behalf of decisionally incapable patients? 1.2a. We suggest ICUs develop resources for educating substitute decision-makers on their role in making decisions on behalf of decisionally-incapable patients. Conditional recommendation, low certainty evidence 1.2b. ICUs should have a standardized process for identifying and documenting the legal surrogate decision-maker for decisionally incapable patients, including those for whom a surrogate cannot be identified, in accordance with local laws and organizational policy. Good practice statement Remark: The process for doing this is best done a hospital/system level and the persons responsible may vary between ICUs but could include the clinical team, social workers, ethics team, legal services, or risk management. 1.3. Should we recommend consultation with such services as clinical ethics, moral distress, palliative medicine, psychology, and/or pastoral/spiritual care to prevent or mitigate conflicts over treatment decisions at EOL? 1.3. We suggest proactive consultation of Palliative Care/Palliative Medicine and/or Ethics Consult Service, when available, to assist with defining goals of care for ICU patients who may no longer benefit from critical care. Conditional recommendation, low certainty evidence 1.4. Should we recommend specific institutional policies to reduce futile and potentially inappropriate treatments when there are conflicts during EOL decision-making? 1.4. We suggest implementing institutional policies to address conflicts over futile and potentially inappropriate treatments in the ICU. Conditional recommendation, low certainty evidence 2.0. Improving provision of EOL care/symptom management for patients in the ICU 2.1. Should we recommend specific treatments to ensure delivery of effective symptom management for dying ICU patients? 2.1. We suggest using protocolized approaches to withdrawal of life-sustaining treatments and symptom management in the ICU, including assessment and management of symptoms pre-extubation, during weaning, and after extubation. Conditional recommendation, moderate certainty evidence 2.2. Should we recommend specific interventions to address the cultural, spiritual, and family traditions and needs of patients and families in the ICU at the EOL? 2.2a. ICU clinicians should explore and support patient and family cultural, spiritual, and family traditions at the EOL. Good practice statement 2.2b. We suggest using a semi-structured approach to supporting patients and families and addressing spiritual care needs including an introductory meeting, weekly follow-up, and post-hospital discharge follow-up. Conditional recommendation, low certainty evidence 2.3. Should we recommend specific interventions such as collaboration with palliative medicine, mental health specialists, spiritual care, or other care process strategies such as care protocols to address, prevent, or mitigate suffering for ICU patients who are at risk of dying in the ICU? 2.3. We suggest consultation and collaboration with an Ethics Consult Service and/or Palliative Medicine, when available, to address the suffering of ICU patients and families at the EOL, when there are challenges in mitigating conflicts, distress, or suffering. Conditional recommendation, low certainty evidence 3.0. Addressing educational needs of clinicians, patients, families, and surrogate decision-makers in ICU EOL care 3.1. Should we recommend strategies and policies to identify, prevent, and mitigate inequities based upon gender, gender identity, sexual orientation, race, ethnicity, faith traditions, country of origin, or socioeconomic status at the EOL? 3.1a. We have insufficient evidence to recommend for or against specific interventions to identify and reduce unmet palliative care needs of specific populations receiving EOL care in the ICU. No recommendation, low certainty evidence 3.1b. ICU clinicians providing EOL care should explore and address patient palliative care needs, considering a patient’s gender, gender identity, sexual identity, race, ethnicity, faith traditions, country of origin, primary language, and socioeconomic status. Good practice statement 3.2. Should we recommend specific interventions or strategies, such as palliative care education, to increase the capability of ICU team members to deliver high-quality EOL care in the ICU? 3.2. We suggest providing education and training in palliative care for all ICU team members to improve the capability of providing EOL care in the ICU. Conditional recommendation, very low certainty evidence 3.3. Should we recommend specific interventions or strategies, such as education efforts, to improve the understanding and expectations of patients at risk of ICU admission, their families, and non-ICU clinicians about EOL care in the ICU? 3.3. We suggest clinicians provide educational interventions to patients/families/surrogates at risk of ICU admission to improve their understanding of ICU and EOL treatment options, realistic treatment outcomes, and advance care planning. Conditional recommendation, low certainty evidence EOL = end of life. Several new topics are addressed in this guideline. For example, PICO question 1.2 includes recommendations for specific interventions to ensure procedural due process for substitute decision makers. We suggest that ICUs develop resources for educating substitute decision makers on their role (Conditional recommendation, low certainty evidence). ICUs should have a standardized process for identifying and documenting legal substitute decision makers, or actions to take when a surrogate cannot be identified, in accordance with local laws and organizational policy (Good Practice Statement). We remark that this process is best done at a hospital/system level and the persons responsible may vary between ICUs but could include the clinical team, social workers, clinical ethics team, legal services, or risk management. In PICO question 1.4, we suggest that healthcare institutions have policies to reduce potentially inappropriate treatments when conflicts arise during EOL decision-making (Conditional recommendation, low certainty evidence). Conflicts during EOL decision-making in the ICU can lead to potentially inappropriate treatments, increase suffering for patients and families, distress clinicians, and disrupt communication between and among patients, families, and treatment teams. In PICO question 2.1, we suggest using protocolized approaches to withdrawing life-sustaining treatment to improve symptom management for ICU patients, including assessment and management of symptoms pre-extubation, during weaning, and after extubation. Using a protocol does not appear to prolong the dying process and uses less medication overall (Conditional recommendation, moderate certainty evidence). In PICO question 3.2, we suggest education and training in palliative care for all ICU team members to improve their capability of providing EOL care in the ICU. Variations in the quality of ICU primary palliative care and readiness to engage in specialty palliative care services significantly influence the use of aggressive interventions at EOL and length of stay. Communication training for ICU teams by palliative care experts reduced the use of renal replacement therapy in patients with poor prognosis (3). A palliative care education intervention in the ICU showed a reduction in ICU length of stay but no impact on patient- or family-centered outcomes (4) (Conditional recommendation, very low certainty evidence). In PICO question 3.3, we suggest that clinicians provide educational interventions to patients at risk of ICU admission, their substitute decision makers, family members, and their non-ICU treatment teams to improve their understanding of ICU and EOL treatment options, realistic treatment outcomes, and advance care planning (Conditional recommendation, low certainty evidence). Patients with serious acute or chronic illness are at risk for ICU admission, which may or may not be congruent with their values and preferences for care. Interventions to educate and assist with decision-making in the inpatient setting, before ICU care is needed, may help to limit unwanted treatments in the ICU and allow more effective symptom palliation at the EOL (5,6). We conclude that expertise supported by ongoing research in EOL care in ICUs is essential for the healthcare team, promotes holistic patient management, and enhances interprofessional communication as well as communication between the care team and patients/families. It supports ethical practices and family needs thus fostering team cohesion and mitigating patient suffering at the EOL.